by RICK BEACH, COPA® SAFETY CHAIR

This unusual CAPS deployment sequence prompted an extraordinary accident investigation. In addition to the pilot actions during the emergency, investigators studied the components of the CAPS system looking to explain an anomaly. This article appears in the January/February 2019 issue of COPA Pilot magazine.

On July 18, 2016, during a C2A formation training event, Cirrus SR22 N678Z experienced a loss of power event and deployed CAPS. The impact attitude after parachute deployment was nose-low, suggesting a low-altitude activation (see Figure 1). The three occupants all survived, although they were walking wounded.

Figure 1: Cirrus SR22 N678Z impacted the ground after a CAPS deployment due to a loss of power event. Three occupants survived the accident that became a fascinating accident investigation of a CAPS activation anomaly.

The resulting accident investigation determined that numerous factors affected the survivable outcome: the loss of power could not be replicated, suggesting an operational issue; the pilot’s decision to deploy CAPS was delayed by his prior training and habits; the activation was estimated at below 500 feet AGL; and the activation sequence appeared highly unusual.

The NTSB investigation invited COPA® to participate as a party representative. As a result, I spent two years working intimately with the NTSB, Cirrus Aircraft and Continental Motors investigators, as well as the pilots involved in the C2A formation training event.

The investigation acquired unique photographic evidence that complemented other data and components recovered from the accident scene. However, not all of the CAPS components could be found, resulting in some complications in the analysis and determination of a probable cause statement.

While the NTSB has released a final report, the investigation team has requested revisions to the analysis and probable cause statements, which are currently in process.

How did the accident unfold?

This CAPS deployment at Colorado Springs happened during a C2A formation training clinic. C2A flight briefings include how to handle all of the planned formation maneuvers, as well as any emergency situations.

In this formation event, there were two aircraft and initially the accident aircraft flew as wingman. Each aircraft had three occupants: the pilot under formation training, a safety pilot and an observer. In the lead aircraft, a professional photographer flew as an observer to document the flight.

One salient point about briefing for emergencies suggested that an aircraft in distress would immediately assume lead, similar to the sobriquet “If it bleeds, it leads.”

Upon experiencing a loss of power and difficulty maintaining altitude, the accident airplane switched positions to assume the lead then reported to ATC that they were returning to the airport about 8 miles away.

Soon it became apparent that the accident airplane would not make the airport despite having the throttle full forward. The pilot began searching for places to land the plane. The safety pilot suggested activating CAPS, but the pilot’s habit persisted in looking for an off-airport landing site. The safety pilot then yelled “Pull CAPS!” so loudly that the back-seat observer could hear it without a headset! The CAPS activation fully deployed the parachute, although the ground impact had a nose-low attitude.

All three occupants egressed the airplane (see Figure 2) and managed to pull the canopy down to the ground and waited for first responders to arrive. Two had minor injuries and one was hospitalized with a transverse process fracture of a lumbar vertebrae, although two days later he flew himself back to Seattle and resumed his airline flying duties.

Figure 2: Immediate aftermath of the CAPS landing showing two of the occupants using their phones in a photo taken by the third. 

What evidence was available?

The wreckage of the Cirrus SR22 was recovered from the prairie and impounded in a secure facility for examination. The engine was dismounted and sent to Continental Motors for a forensic examination.

Interviews with the occupants of both airplanes in the formation flight reported the sequence of events including a delay in activation and then the unusual deployment of the parachute canopy.

Recorded engine data from the Avidyne MFD provided information about the accident flight and several earlier flights. The NTSB Vehicle Recorder Lab analyzed those data files and produced a typical report. Unfortunately, the data did not include altitude information to determine height of the CAPS activation.

Uniquely, the investigation team was given both a GPS flight track from the lead airplane and a photographic archive covering the activation, deployment and recovery of the accident airplane.

Mike Radomsky provided a CloudAhoy flight track based on a Bad Elf® GPS device. That flight track followed the accident plane (see Figure 3) in its gradual descent (purple) until it deployed CAPS then circled for about 20-30 minutes (yellow) until first responders arrived.

Figure 3: Flight track of the second airplane in the two-ship formation that followed the accident airplane (purple) during its gradual descent and CAPS deployment then orbited the downed airplane (yellow) until the arrival of first responders.

Greg “Baron von Speed” Akers provided an extensive photo archive of 161 high-resolution images taken from the second airplane. The raw images contained 8688 x 5792 pixels that were very useful for detailed examination.

Metadata in these photos included a time stamp to a hundredth of a second. Using that time stamp, we could interpolate the closest GPS location and altitude of the camera from the Bad Elf data file.

Engine analysis and testing

The MFD engine data file contains both the accident flight and several prior flights. These were examined and analyzed looking for evidence of a loss of power event.

The data analysis found that the oil pressure was low, ranging from 32 to 37 psi during the last 10 minutes of the flight. The fuel flow ranged from 7 to 15 gph during that time. Engine rpm ranged from 2,000 to 2,600 rpm with it slowly declining in the last two minutes of flight.

The IO-550N engine was equipped with an after-market supercharger by Forced Airmotive Technologies. The engine test conducted by Continental Motors found that the engine ran without issues on the test stand at sea level in Mobile, Alabama.

None of these tests and analyses provided a simple answer to the loss of power event. High density-altitude conditions existed during the summer flight at the elevation of Colorado Springs. From my vantage point, without any experience in operating a supercharged engine, nor flying in a wingman position in a formation flight, I wondered if the low fuel flow might have contributed to the loss of power at high rpm.

About that anomalous CAPS deployment sequence

From the earliest opportunity to view the photo archive, one photo stood out. Figure 4 shows the parachute bag (white) drooping well below the aircraft. None of us had ever seen that in a rocket-assisted CAPS deployment sequence where the parachute is extracted high above the plane to inflate in the slipstream.

Figure 4: What?! This photo shows the packed parachute bag drooping below and behind the aircraft during the CAPS deployment sequence.

Ultimately, the parachute did inflate somewhat normally. However, given the time stamps of the photos, inflation of the canopy took a few seconds longer than usual. Add the reality of a low-altitude activation and the ground impact in a nose-low attitude would be expected.

But how low? 

Photo analysis by COPA

As the COPA representative, I prepared an analysis report of the photo archive to estimate the altitude of the CAPS activation. Coincidentally, the initial photos included background details of the three radar domes on the ground at Schriever Air Force base near Colorado Springs (see Figure 5).

Figure 5: Detail from one of the first photos in the deployment sequence shows the radar domes at Schriever Air Force Base. Using the altitude of the camera and the location of the radar domes, we could estimate the altitude of CAPS activation as 472 feet AGL. (Due to the circumstances of the event, some photos had camera shake.)

Using those radar domes and the data from the camera, we could triangulate the position and altitude of the camera with the radar domes. Then by estimating the position of the accident aircraft, we could compute the altitude of the CAPS deployment as approximately 472 feet AGL immediately after activation.

Collecting components of the CAPS system

Typical of any accident investigation, the four corners of the aircraft are located. With the wreckage mostly intact, that was easy.

For a CAPS deployment, the investigators want to collect all of the components of the CAPS system that separate from the airplane: CAPS cover, rocket, lanyard, incremental bridle and parachute deployment bag.

The first responders found the white parachute deployment bag. Although they didn’t mark where it was found, examining the photo archive revealed that it was only a few hundred feet from the wreckage.

However, that parachute bag was missing a few key components. Normally, the bag is connected to the rocket using a yellow Kevlar incremental bridle and a wire lanyard that goes to the pickup collar that fits around the rocket. The incremental bridle was attached to the bag but the wire lanyard was broken, with the rocket and pickup collar missing. We wanted to find the rocket and pickup collar, as well as the CAPS cover.

To find the CAPS components on the flat grassy prairie, we combined both a foot-search (the owner refused permission to drive on the prairie to avoid damaging the grass) and a photo search using a professional drone operator found through COPA connections.

The locations of the other components could be much farther away, especially the rocket. We estimated that the search area might be a trapezoid up to 1.5 miles along the route of flight and extending up to 1 mile wide.

Some items were easy to locate. The white CAPS cover is big enough that the drone found it visually and the team recovered it quickly (see Figure 6). The other missing components were a bit more challenging. One item caught my eye as we were walking our search path – the Teflon cable sheath from the wire lanyard between the rocket and the incremental bridle. It was just lying on the ground.

Figure 6: The CAPS cover was big enough and white enough to be easily found.

Drone search for the rocket

The key missing piece was the rocket – a red aluminum cylinder about 8 inches long and 3 inches in diameter, likely to be found anywhere in the specified trapezoid from the wreckage site.

Ultimately, we obtained 3,500 photos and used an image-processing firm to look for clusters of red pixels. Despite having numerous targets that found red fence posts and even a Coke can, there was no rocket.

Fractured CAPS lanyard examined by NTSB Materials Lab

The key mystery from the anomalous CAPS deployment by N678Z was the broken wire lanyard. We didn’t have the rocket or the pickup collar but we did have two pieces of the wire cable that normally looped around the rocket.

The NTSB Materials Lab investigated the broken wire and compared it to test samples provided by Cirrus Aircraft. They used a scanning electron microscope to examine the properties in the fracture and the metal deposits made from the pickup color assembly. These tests determined that the fracture locations were consistent with the small radius support holes in the pickup collar (see left item in Figure 7).

Figure 7: Two versions of the CAPS pickup collar that attaches to the rocket and carries the lanyard through the support holes and around the collar. Left part is the original design and the right part is the design introduced with the 3600-pound gross weight parachute, also used in all electronic ignition CAPS assemblies since 2013.

Cirrus Aircraft had previously modified the design of the pickup collar to accommodate the heavier parachute in the 3,600-pound gross weight SR22/SR22T models (see right item in Figure 7). The improved part has a larger radius support for the wire lanyard and is used in all electronic ignition CAPS assemblies.

What do I think happened?

There seems to be two things that happened coincidentally in this accident:

1.     The pilot’s delayed activation of CAPS resulted in a low-altitude deployment below 500 feet AGL providing insufficient time to land level under canopy.

2.     The CAPS rocket fractured the wire lanyard during extraction of the parachute deployment bag resulting in a delayed (by only a couple of seconds) inflation of the parachute canopy.

None of the photographs show the rocket. The first image with the activation sequence shows the CAPS cover (shown extreme left of Figure 8). Upon close examination, one sees the yellow incremental bridle flopping behind the plane, as if it was recoiling from the broken lanyard.

Figure 8: Obscure first photograph shows the CAPS cover off to the far left in the image. Unfortunately, the photographer was surprised by the pilot’s delay to activate CAPS, so no images of the rocket were captured.

The next image suggests that the parachute bag has plopped onto the horizontal stabilizer. Subsequent images show the parachute bag dropping behind and below the plane, eventually releasing the canopy that inflated normally.

Safety Recommendations to NTSB

As a party to the investigation, I submitted the following safety recommendations for their consideration:

1.     Use the final report to inform Cirrus pilots that Cirrus Aircraft has redesigned the rocket pickup collar.

2.     Consider habit interference that affected the pilot’s decision to activate the CAPS parachute system.

3.     Recognize the distraction and stress from the formation training clinic scenario.

Redesigned rocket pickup collar

Any Cirrus airplane with the electronic ignition assembly already contains the revised design of the pickup collar (see Figure 7). That new design alleviates the stress on the lanyard and reduces the possibility of a similar fracture.

By the end of 2018, Cirrus Aircraft reports that 75 percent of the Cirrus fleet incorporates this updated design. Due to the 10-year repack schedule, the remaining aircraft will be retrofitted from 2019 to 2023.

Habit interference

Using the NTSB Aviation Coding Manual for identifying factors in an accident, I suggested adding the human factors finding “31000 Psychological Condition / 31270 Habit Interference.”

In the public investigation docket, the back-seat observer includes this detail:

[The pilot] appeared to be looking for a suitable landing site in the directions of Springs East and Bullseye airports. [The pilot] then reached up and removed the CAPS handle cover and tilted the handle down still looking around the area. I again looked at the MFD and saw 56 percent power with the throttle full forward. [The safety pilot] then yelled, “Pull it now!” [The pilot] then pulled the handle.

Subsequently, the safety pilot provided additional detail that was included in the factual report:

The safety pilot reported that the pilot in command would pull the CAPS unless incapacitated, as briefed during preflight briefing. According to the safety pilot, the pilot’s previous and overriding training habit kicked in as he looked for a place to land. The pilot verified with the safety pilot that he intended to deploy the CAPS and pulled the CAPS handle at the safety pilot’s second request.

That reference to habit suggests an important pilot-training aspect of this investigation: overcoming habits learned in prior flying through a suitable transition training program to flying a CAPS-equipped airplane.

Self-induced pressure

To help pilot training, another finding from human factors would be the addition of “31000 Psychological Condition / 31201 Self-Induced Pressure.” Kudos to the organizers of the C2A formation flying clinics who subsequently expanded their preflight briefings to acknowledge the additional pressure of learning to fly in a formation.

What You Can Do

This accident investigation involved several interesting paths that come back to how a Cirrus pilot handles an in-flight emergency.

Have you developed the habit of “Consider CAPS”?

We recommend that every Cirrus pilot develop habits that promote the use of CAPS during an emergency. Add a first step to each emergency procedure:

Step 0. CAPS … Consider

As well, we recommend using the departure briefing found in the Flight Operations Manual. Add the practice of reaching up and grasping (but not pulling) the CAPS handle when you climb to the CAPS-viable altitude for your plane. This helps develop muscle memory that may make consideration of CAPS more likely in your emergency.

Stress and distraction

A possible factor in this accident was the additional stress and distraction of flying a formation training event. Would you recognize the extra energy and nervousness involved in any training mission? How would you mitigate that stress and distraction?

High altitude operations are different

Another aspect of this accident involves flight in an area of hot and high conditions. The density altitude was several thousand feet higher than the terrain elevation. Consequently, engine operations need to be different, consistent with your engine model, whether normally aspirated or boosted with a turbo or supercharger. Familiarize yourself with how best to handle hot and high conditions.

Rick Beach joined COPA in 2001 and bought his Cirrus SR22 because of the community of owners and their focus on safety. He regularly posts online as the COPA de facto “accident historian” and is the COPA Safety Chair. In his spare time, he works with educational non-profits to improve K-12 math and science education.

Comments are closed on this blog. Discussion of the accident can be viewed by COPA members here: CAPS Save #67 in SR22 N678Z east of Colorado Springs, CO [18 June 2016]